A Flexible Smoother Adapted to Censored Data With Outliers and Its Application to SARS-CoV-2 Monitoring in Wastewater

A sentinel network, Obépine , has been designed to monitor SARS-CoV-2 viral load in wastewaters arriving at wastewater treatment plants (WWTPs) France as an indirect macro-epidemiological parameter. The sources of uncertainty such a monitoring system are numerous, and the concentration measurements it provides left-censored contain outliers, which biases results usual smoothing methods. Hence, need for adapted pre-processing order evaluate real daily amount viruses each WWTP. We propose method based on auto-regressive model censored data with outliers. Inference prediction produced via discretized smoother makes very flexible tool. This is both validated simulations from . resulting smoothed signal shows good correlation other epidemiological indicators currently used by provide estimate virus circulation over watersheds corresponding about 200 WWTPs.